Advising Women at High Risk of Breast Cancer
Advising Women at High Risk of Breast Cancer
ABSTRACT: Women with any family history of breast cancer assume a high probability of risk. Counseling women involves ascertainment of an accurate family history and use of the best predictive models to assess both the risk of a known mutation and the risk of breast cancer. This risk must then be considered in the contexts of both the woman’s lifetime and the next decade, in each instance carefully separating the risk of developing cancer from the risk of mortality. These two risks are often emotionally melded in women who have watched a loved one die of cancer. The options for a woman at significantly increased risk of breast cancer include optimal surveillance, chemoprevention, and prophylactic surgery. This entire field is in continuing evolution as better methods of diagnosis, screening, and chemoprevention continue to enter clinical practice.
Awareness of inherited predilections for the development of breast cancer has been coupled with an information explosion in the lay literature about the subject of breast cancer screening techniques and methods of therapy, creating an anxious public. In the last decade, BRCA1 and BRCA2 were identified as high-risk mutation genes in relationship to breast cancer. PTEN mutation associated with Cowden syndrome is also clearly related to a moderate risk for breast cancer. It was anticipated that a host of other genetic mutations related to high or moderate penetrance breast cancer association in families would follow. To date, no other dramatic breast cancer-related gene has been identified, despite many other families with an apparent pattern of dominant transmission (Table 1).
Women with any family history of breast cancer assume a very high risk of developing the disease themselves. Counseling such women begins with a three-or four-generation family history. With a large family, such a history may be quite informative; with sparse families, it is less so. The Gail model has been validated as a reasonable instrument for assessing population risk, but it is of limited value in hereditary cancer families. The Claus model is more accurate in families at high risk but underestimates paternal transmission. Two models that are specifically directed toward the identification of high-risk families are available on the Internet from Myriad Genetics[ 2] and Johns Hopkins University (BRCAPRO).
Three Patient Groups
In clinical consultation, three populations of women present with concerns about high risk of breast cancer. The first group are those who have had genetic testing and have been identified as expressing a mutated gene associated with high-penetrance breast cancer. Although we are unable to assign a risk specific to any individual, this population of women has a clearly defined risk of breast cancer development between 55% and 85% in the case of BRCA1 and BRCA2 mutation carriers, depending somewhat upon the specific mutation. These mutations are not only associated with high risk, they are associated with a much earlier risk of breast cancer than is seen in the sporadic pattern.
The second population comprises women who present with a strong family history suggestive of inheritable breast cancer but test negative for BRCA1 and BRCA2. They are clearly at increased risk, but it is impossible at present to quantitate the magnitude of this increase.
The third population are those who have a strong family history but have not been tested. This last group is of particular concern when they seek a major intervention (eg, prophylactic mastectomy). If a family member with early breast or ovarian cancer is alive and can be tested for the presence of a gene mutation, such testing should always be encouraged. Such a mutation has only a 50% likelihood of being inherited by the proband in consultation. Being able to identify which woman has not inherited the mutated gene and has only a normal population risk of breast cancer spares not only the possibility of prophylactic surgery but great anxiety, both at present and in the years to come.
Three Major Questions
Three major questions dominate a consultation with a woman who is not aware of having breast cancer but is concerned about her family history.
The first question focuses on the best method of surveillance in order to detect breast cancer, should it arise, as early as possible. Several recent studies suggest that magnetic resonance imaging is better than screening mammography at the early detection of breast cancer in young women who are at high risk.[4,5] The concordance of these published trials gives more weight to this finding than the numbers or statistical strength of any of the studies, singly or additively.
Whether magnetic resonance imaging or, indeed, mammography should be performed more often than annually and at what age it should be initiated have not been clearly answered. The old recommendation of Dr. Henry Lynch that 10 years prior to the age at which a mother or sibling developed breast cancer is when to start radiographic screening appears prescient as anecdotal and series data continue to accumulate.
Clinical breast examination remains important, as many cancers in this sensitized population continue to be detected by patients themselves. Although it is not clear that earlier patient detection affects mortality, it may limit the mode and extent of therapy required for treatment.
A final important aspect of surveillance involves the immediate biopsy of suspicious lesions. This can be accomplished by fine-needle aspiration on the day that a lesion is first identified, rather than waiting a month or two to see if it is a physiologic abnormality that will regress spontaneously. The latter approach is to be decried in these women at increased risk. The role of real-time ultrasound in screening this high-risk population remains uncertain.
Although surveillance is very important in women at increased risk, its sensitivity clearly varies with the individual. Whether breasts are soft, hard or nodular, radiolucent or mammographically dense, of average size vs very large or very small, and if the subject is compliant vs noncompliant with a regular screening program are all factors that clearly influence the efficacy of surveillance.
No prospective, well-defined surveillance program in a genetically defined population has clearly delineated the role of a specific screening protocol in this population. A series of small studies are under way, and interim reports are now appearing. Nevertheless, previous retrospective evaluations of surveillance outcome in "high-risk" women (variously defined) do raise alarm by showing that from 20% to more than 30% of the patients in such programs have lymph node involvement at the time of tumor detection.
Approaches to Prevention?
The second major question that arises is concerned with the prevention of breast cancer in women at high risk. Lifestyle changes have often been instituted by these women themselves. Some become vegetarians, others increase their athletic activities, and some adapt techniques alleged to diminish stress and tension. At present, there are no data to suggest that any of these approaches diminishes the risk of breast cancer or the rate of its progression.
Activity has been associated with a lower incidence of breast carcinoma in the general population, but only at the extremes. Women who have an extremely sedentary lifestyle and are overweight have an increased risk of breast cancer, and those who exercise- particularly to the point of ovarian suppression (eg, marathon runners)-have a diminished risk of breast cancer. Such data are not available in mutation-positive high-risk women.
The only dietary factor that appears repeatedly in cohort studies as a risk modifier is the use of alcohol. One glass of wine daily produces no demonstrable increased risk. Two glasses are associated with a 20% to 30% increase in baseline risk, and three or more glasses, with an increase of roughly 50%. Whether these increases apply to the population of genetically defined high-risk women is unproven.
• Hormonal Therapy—Tamoxifen has been shown to prevent contralateral breast cancer by roughly 50% when taken for 5 years.[7,8] The limited data available suggest that this is true in BRCA2 mutation carriers, but it is not clear that it is true in BRCA1 patients. This distinction has arisen because the majority of sporadically arising breast cancers are hormone receptor-positive, as are most BRCA2-related breast cancers. A minority of BRCA1 mutation breast cancers are hormone receptor-positive. The fact that oophorectomy for women with hereditary breast/ovarian syndrome (BRCA1/BRCA2) appears to prevent breast cancer in BRCA1 as well as BRCA2 mutation carriers suggests that more data are needed before tamoxifen can be considered ineffective in the BRCA1 population.
The current Study of Tamoxifen and Raloxifene (STAR) trial is comparing tamoxifen with raloxifene (Evista) to prevent breast cancer. European trials are evaluating the efficacy of aromatase inhibitors in this setting. Early data from the Arimidex, Tamoxifen Alone or in Combination (ATAC) trial suggesting that anastrozole (Arimidex) reduces the incidence of contralateral breast cancer even more effectively than does tamoxifen have been encouraging in this respect. The actual utility of these alternative agents outside of a clinical trial remains to be established.
• Oophorectomy—At least three reports of oophorectomy in BRCA1 and BRCA2 mutation carriers are associated with a reduction to half the rate of subsequent breast cancer development.[ 9,10] Although the numbers in these trials are small, the effect is consistent among them. This is particularly appealing prophylactic surgery because the risk of ovarian cancer causes such concern in women with BRCA1 or BRCA2 mutations. Although the risk of ovarian cancer is lower than that of breast cancer, our ability to detect it at an early stage is almost nonexistent. The possibility of proteomic diagnosis in the near future would be a bright hope to that end.